Chromatography of metabolites in plasma and urine following oral administration of anthocyanin rich capsules

(1)

Faculty of Science and Technology

MASTER’S THESIS

Study program/ Specialization:

Biologisk kjemi

Spring semester, 2010

Open / Restricted access Writer:

Christina Helén Nilsen ………

(Writer’s signature) Faculty supervisor: Kåre B. Jørgensen

External supervisor(s): Grete Jonsson

Title of thesis:

“Chromatography of metabolites in plasma and urine following oral administration of anthocyanin rich capsules”

Credits (ECTS): 60 studiepoeng Key words:

Anthocyanins Benzoic acids GCMS LLE Metabolism

Pages: 72

+ enclosure: 11

Stavanger, 14.06.2010 Date/year

(2)

Chromatography of metabolites in plasma and urine following oral administration of

anthocyanin rich capsules

Christina Helén Nilsen

Master thesis: Biological Chemistry

Department of Mathematics and Natural Science Spring 2010

(3)

Abstract

Anthocyanins (ACs) are powerful antioxidants widely distributed in fruits and vegetables.

Several international studies suggest that anthocyanins have positive effects on the health, including various chronic diseases. To be able to study the possible anthocyanin mechanisms in the human body, it is of importance to know their metabolism. MEDOX® is an AC rich product made from bilberries and blackcurrant and consists of 17 different anthocyanins with delphinidin-3-O-β-glucopyranoside and cyanidin-3-O-β-glucopyranoside as the main

constituents. There are only 5 different aglycone structures of these 17 anthocyanins;

delphinidin, cyanidin, peonidin, petunidin and malvidin. An extensive metabolism of ACs is indicated following oral administration of this supplement. Free and conjugated benzoic acids (BAs) with functional groups corresponding to the anthocyanin B-ring structure have been recognized as metabolites in urine and plasma. Correspondingly, gallic acid, protochatechuic acids, vanillic acid, syringic acid and 3, 4-dihydroxy-5-methoxybenzoic acid have been suggested as metabolites of delphinidin, cyanidin, peonidin, malvidin and petunidin respectively. Hydroxy benzoic acids are relatively polar compounds, and derivatization is therefore usually necessary prior to gas chromatography mass spectrometry (GCMS) analysis.

The objective of this thesis was to develop a robust analytical method for determination of these BAs in addition to 4-hydroxybenzoic acid which is a suggested metabolite of the AC pelargonidine. Liquid-liquid extraction (LLE) was chosen as the best sample extraction method as compared to solid phase extraction (SPE) and solid phase analytical derivatization (SPAD) with respect to recovery, reproducibility and sample purity. LLE was also less time consuming than SPE. The recoveries of BAs in urine ranged from 62 – 121 %, and recoveries in plasma from 37 – 158 %. All BAs were identified and quantified in urine and plasma after oral administration of MEDOX® in increased levels compared to before intake. All

metabolites were detected as both free BAs and in conjugated form linked with glucuronic acid, in both urine and plasma. Results obtained from the work of this thesis suggest that the ACs in this product metabolize to their corresponding BAs.

(4)

Acknowledgements

First I would like to gratefully acknowledge my supervisor Grete Jonsson for her guidance, help and support during the work of this thesis. Your teaching philosophy has made me constantly challenge myself and I have learned a lot from you. I appreciate the time I have had at SUS and what I have been able to experience during my work there. I am also very grateful to Atle Nævdal at IRIS who let me use the GCMS instrument very flexibly and set aside time to give me technical support and help. Thanks to Cato Brede for enthusiastically offering advice and technical support. A big “thank you” goes out to the nice ladies in the lab next door for their help by drawing blood and analyzing creatinine. I would also like to thank my University supervisor Kåre Jørgensen for technical and practical help with my thesis. I am thankful to my family for always showing their interest in what I do and for being proud of me. Finally I would like to thank my fiancé Øyvind for his support, encouragement and motivation.

Christina Helén Nilsen, Stavanger 2010

(5)

Abbreviations

3OMGA - 3-O-methylgallic acid

4OMGA - 4-O-methylgallic acid

4CL - 4-coumarate-CoA ligase

AA - Amino acid

AC - Anthocyanin

ACCase - Acetyl CoA carboxylase

AcN - Acetonitrile

ACT - Acyltransferase

ADP - Adenosine diphosphate ANS - Anthocyanidin synthase ATP - Adenosine triphosphate

BA - Benzoic acid

BSA - N,O-bis(trimethylsilyl) acetamide BSTFA - N,O- bis(trimethylsilyl) trifluoroacetamide C3G - Cyanidin-3-O-β-glucopyranoside

C4H - Cinnamate 4-hydroxylase CE - Capillary electrophoresis

CHI - Chalcone isomerase

CHI - Chalcone synthase

CI - Chemical ionization

CoA - Coenzyme A

Conc. - Concentration

D3G - Delphinidin-3-O-β-glucopyranoside DFR - Dihydroflavonol 4-reductase d-GA - Deuterated gallic acid (GA)

d-HBA - Deuterated 4-hydroxybenzoic acid (HBA) DHMBA - 3,4-dihydroxy-5-methoxybenzoic acid DR - Derivatization reagent

EDTA - Ethylenediaminetetraacetic acid

EI - Electron ionization

EtAc - Ethyl acetate

Eq. - Equation

F3H - Flavonone 3-hydroxylase F3´H - Flavonoid 3´-hydroxylase F3´5´H - Flavonoid 3´, 5´-hydroxylase

FI - Field ionization

Fig. - Figure

GA - Gallic acid

GC - Gas chromatography

GCMS - Gas chromatography mass spectrometry GLC - Gas-liquid chromatography

GSC - Gas-solid chromatography

GT - Glycosyltransferase

HBA - 4-hydroxybenzoic acid HCl - Hydrochloric acid

(8)

HMBA - 4-(hydroxymethyl) benzoic acid HMDS - hexamethyldisilazane

HPLC - High performance liquid chromatography Ill. - Illustration

IS - Internal standard

LLE - Liquid liquid extraction LOD - Limit of detection

LOQ - Limit of quantification m/z - Mass to charge ratio MAT - Malonyltransferase

MeOH - Methanol

MPI - Multi photon ionization

MS - Mass spectrometry

MTBE - Methyl tert-butyl ether

OH - Hydroxy-group

P - Phosphate

PAL - Phenylalanine ammonia-lyase PCA - Protocatechuic acid

PEP - Phosphoenol pyruvate

PI - Photo ionization

PP - Polypropylene

QP - Quadrupole

RIC - Reconstructed ion chromatogram

RP - Reversed phase

S/N - Signal-to-noise ratio

SA - Syringic acid

SPAD - Solid phase analytical derivatization SPE - Solid phase extraction

Std. - Standard

Tab. - Table

TIC - Total ion current TMCS - Trimethylchlorosilane

TMS - Trimethylsilyl

TMSI - N-trimethylsilylimidazole

UV - Ultra violet

UV/Vis - Ultra violet/visible

VA - Vanillic acid

(9)

1. Introduction

Flavonoids are a common term of numerous polyphenolic compounds found ubiquitously in nature. Flavonoids are plant pigments giving color to certain fruits and flowers. The perhaps most intriguing aspect of flavonoids are their antioxidant function which holds a potential health benefit for humans. The main dietary sources of these antioxidants for humans are fruits, vegetables and various beverages. Based on their chemical structure flavonoids are categorized into ^[1-3]:

- Flavones - Flavanones - Flavonols - Isoflavones - Catechins - Anthocyanidins - Chalcones

Anthocyanins (Fig. 1) are a part of the vast flavonoid family, consisting of an aglycone (anthocyanidin) part and a glycone (sugar) part. They are a group of naturally occurring compounds which are powerful antioxidants. The anthocyanins (ACs) are responsible for the color of many fruits, vegetables, and are also found in grain products. As a result, the ACs participate in a typical everyday life as a part of the human diet. The intake of anthocyanins in humans has been estimated to be 180-215 mg/day, based on numbers from the United States of America. This estimate is considerably higher than the intake of other flavonoids, including quercetin, kaempferol, myricetin, apigenin and luteolin, which is approximately 23 mg/day ^[4,

5].

Figure 1. Anthocyanin structure

ACs are perhaps best known as color giving plant pigments, however they have other more important features. In plants anthocyanins contribute to the immune system protecting them from harmful ultraviolet (UV) rays, and play an important part in the pollination process ^[6]. The main objective for the interest in ACs considering humans is the alleged health benefits due to their antioxidant effect. Because of the possibility of ACs being source of several health aspects, they have been the target of a number of research projects and studies ^{[7, 8]}. In this thesis, potential metabolites of the anthocyanins containing the six most prominent aglycone structures have been studied (Table 1). Five of these aglycone structures with different sugars attached, are found in the blueberry and blackcurrant based product MEDOX® which has been used as supplement for oral administration to study a possible

(10)

metabolism of the anthocyanins. The sixth anthocyanin, pelargonidin, should not be found in this product. The R5, R6 and R7 groups are the same for all these compounds: OH, H and OH respectively ^[3].

Table 1. Most common aglycone structures Aglycone R5´ R3´

Delphinidin OH OH

Cyanidin OH H

Peonidin OCH3 H Petunidin OH OCH3

Malvidin OCH3 OCH3

Pelargonidin H H

Benzoic acids (BAs) consist of a benzene ring directly bonded to a carboxyl group (Fig. 2).

BAs can also be called phenolic acids in some cases and vice versa, depending on the

structure. Benzoic acids always have the acidic group bound directly to the benzene ring; this is not always true for phenolic acids. Phenolic acids have a hydroxy (OH) group bound directly to the benzene ring, which is not always the case for benzoic acids. The compounds sharing these qualities and can be referred to as both, are also often called hydroxy-benzoic acids ^[9].

Figure 2. Benzoic acid structure

Based on research on anthocyanins and metabolic research of ACs, benzoic acids are suggested to be metabolites of anthocyanins. The BA metabolite’s functional groups correspond to the anthocyanin B-ring structure. These benzoic acids can be found either as free or conjugated acids, where in the latter case deconjugation might be necessary for a quantitative analysis. The conjugated acids are usually linked to glucuronic acid

(glucuronidated) and/or sulphuric acid. Methylation is also a common metabolic route, allowing some BAs to transform into others ^[10-15].

The benzoic acids analyzed in this thesis is corresponding to the six most common aglycone structures of anthocyanins (Table 2). All the BAs share the qualities qualifying their structures as both benzoic and phenolic acids.

(11)

Table 2. Analyzed benzoic acids and their parent compounds BA

(chemical name)

BA (abbreviation)

BA

(common name) Parent compound 3,4,5‐trihydroxybenzoic acid GA Gallic acid Delphinidin

3,4‐dihydroxybenzoic acid PCA Protocatechuic acid Cyanidin 4‐hydroxy‐3‐methoxybenzoic acid VA Vanillic acid Peonidin 3,4‐dihydroxy‐5‐methoxybenzoic acid DHMBA ‐ Petunidin 4‐hydroxy‐3,5‐methoxybenzoic acid SA Syringic acid Malvidin

4‐hydroxybenzoic acid HBA ‐ Pelargonidin

The structures of the suggested hydroxy-benzoic acid metabolites of the selected anthocyanins are corresponding to the different AC’s B-ring structure (Fig. 3).

Figure 3. Structure of the benzoic acid metabolites analyzed in this thesis

MEDOX® (Fig. 4) is a unique and natural anthocyanin product containing 17 different anthocyanins (Biolink Group AS, Norway).

Figure 4. MEDOX® product

MEDOX® is a product combining bilberries (Vaccinum myrtillus) and blackcurrant (Ribes nigrum). The molecules believed to have the highest biological potential among ACs are the ones possessing a so-called “ortho-dihydroxy phenol” structure. Two of these are cyanidin-3-

(12)

O-β-glucopyranoside (C3G) and delphinidin-3-O-β-glucopyranoside (D3G), which are vastly overrepresented in MEDOX®. In fact C3G and D3G accounts for 91 % of the 80 g AC content in MEDOX® capsules ^{[16, 17]}.

Of the 17 different anthocyanins found in MEDOX®, there are only five different aglycone structures (Table 3).

Table 3. Anthocyanins found in MEDOX®

Aglycone Variations

structure 3‐O‐β‐glucopyranoside 3‐O‐β‐galactopyranoside 3‐O‐α‐arabinopyranoside 3‐O‐β‐rutinoside

Delphinidin x x x x

Cyanidin x x x x

Peonidin x x x

Petunidin x x x

Malvidin x x x

Preparation of samples prior to analysis is necessary to make the sample material compatible with the instrument utilized. Complex matrixes such as urine and plasma are unfit for direct analysis because they contain an excess of biological materials such as protein, salt and cell products, making a purification step essential. By applying an extraction method for purification, the sample is also allowed to be concentrated. Sometimes extraction alone is insufficient and an additional modification is needed, such as derivatization. Derivatization is performed on a sample material to make it suitable for the following analysis ^[18].

Common sample preparation methods for ACs in biological samples are liquid-liquid

extraction (LLE) and solid phase extraction (SPE) following pretreatment of deconjugation in most cases [7, 8, 10, 19-23].

Chromatography is defined as a process for separating compounds in a mixture. The term chromatography is collective for several separation techniques, for which the principle is to separate analytes by distribution over two phases, a mobile phase and a stationary phase. The mobile phase is either a liquid or gas; hence the names liquid chromatography (LC) and gas chromatography (GC) ^[18].

For the separation, detection and analysis including identification and quantification of the compounds in this thesis, capillary gas chromatography coupled with mass spectrometry (GCMS) was chosen. GCMS was considered a good choice due to the techniques separation and identification qualities ^{[24, 25]}. Various separation techniques have been used for separation and identification of phenolic compounds, mainly HPLC and GC [7, 8, 10, 19-23]. Capillary electrophoresis (CE) has also been utilized for separation of phenolic compounds, however it is not common ^[26]. A common advantage when applying techniques such as HPLC and CE, compared to GC, is that these methods do not usually require a derivatization step before quantitative analysis which can be quite time consuming. This argument is contributing to the choice of HPLC as the most commonly used method for qualitative and quantitative analysis of plant phenolic compounds. The detector principles used with HPLC is typically UV, electrochemical and fluorescence [10, 15, 18]. Because the UV/Vis spectrum does not supply sufficient identifying power, and HPLC does often not provide sufficient separating

performance, GCMS is a good choice of method for analysis of phenolic compounds [24, 25, 27].

(13)

The overall aim of the thesis is the development of a robust chromatographic method for determination of selected metabolites in urine and plasma following oral administration of MEDOX®. The selected metabolites are free and glucuronic acid conjugated benzoic acids:

- 3,4,5-trihydroxybenzoic acid (GA) - 3,4-hydroxybenzoic acid (PCA)

- 4-hydroxy-3-methoxybenzoic acid (VA)

- 3,4-dihydroxy-5-methoxybenzoic acid (DHMBA) - 4-hydroxy-3,5-dimethoxybenzoic acid (SA), and:

- 4-hydroxybenzoic acid (HBA)

The first five of these benzoic acids, respectively, are expected metabolites from the MEDOX® supplement; whereas the last acid (HBA) is an expected metabolite from pelargonidine which is an anthocyanin found in strawberries amongst others.

Based on the results acquired by the use of the developed optimized method, there should be possible to discuss whether these benzoic acids really are metabolites of the anthocyanins found in MEDOX®.

(14)

2. Theory

2.1 Anthocyanins

Anthocyanins are of great interest because of their alleged antioxidant activity. They are part of the flavonoid family which are synthesized from phenylalanine and are characterized by a benzo- γ-pyrone structure ^[2]. The skeletal carbon structure (Fig. 5) possessing 15 carbon atoms; two benzene rings joined by a linear three carbon chain, can be described as C6-C3-C6, containing three phenolic rings called A, B and C. The C-ring can also rarely consist of two carbon atoms. Usually the B-ring is in 2 position of the C-ring; however it can also be located at position 3 and 4 ^{[2, 28]}.

Figure 5. General flavonoid structure

Phenols and polyphenols, including flavonoids and ACs, are effective antioxidants since the radical products of these molecules are resonance stabilized and thus relatively stable ^[29]. It is believed that ACs have positive effects on several health aspects also in a preventive way.

The metabolism and bioactivity of these antioxidants are two features of great importance to study to understand their effect.

2.1.1 Biosynthesis and metabolism

The anthocyanin biosynthesis is also often called the flavonoid pathway since other flavonoids are also synthesized by this pathway. The ACs are assembled by two separate biosynthesis paths which result in the amino acid (AA) phenylalanine (Fig. 6) and malonyl- coenzyme A (coA) (Fig. 7) which are both essential in the joined synthesis of anthocyanins ^[3,

6, 29-31].

(15)

Figure 6. Synthesis of the amino acid phenylalanine, the starting point of AC biosynthesis

The abbreviations ATP, ADP and P in the figure above (Fig. 6) stand for adenosine triphosphate, adenosine diphosphate and phosphate respectively. ATP functions as a coenzyme and is often used for energy transfer in the cells for metabolism. ATP is

transformed to ADP by phosphorylation ^{[29, 30]}. The Shikimate pathway leads from PEP and E4P to chorismate. Chorismate is thereafter used as substrate to form the aromatic AAs phenylalanine, tyrosine and tryptophan. Only phenylalanine is involved in the synthesis of flavonoids and ACs. Phenylalanine synthesis from L-Arogenate is the characteristic route in higher plants. Some bacteria form the aromatic AAs directly from prephenate ^{[3, 32]}.

Figure 7. Synthesis of Malonyl-CoA, which is needed for AC synthesis

(16)

Acetyl-CoA carboxylase (ACCase) is the enzyme which transforms Acetyl-CoA into Malonyl-CoA which is used further for AC synthesis ^{[29, 30]}. The two separate streams resulting in phenylalanine and malonyl-CoA meet and are enzymatically joined together by chalcone synthase (CHS), generating a transitional chalcone which is isomerized to the flavonone naringenin. Anthocyanin production happens in plants through a series of enzymatically assisted steps which constitutes the anthocyanin biosynthesis (Fig. 8) ^{[3, 6, 31]}.

Figure 8. Anthocyanin (flavonoid) biosynthesis, abbreviations explained in the text Phenylalanine ammonia-lyase (PAL) is the first enzyme involved in the biosynthetic pathway of ACs. PAL leads to the cleavage of the amino-group of phenylalanine which is released as ammonium which is utilized in glutamine synthetase. Glutamate and glutamine serve as nitrogen donor to chorismate in the pathway leading to phenylalanine. Transcription of the PAL gene is enhanced by light, plant growth regulation as wounding (excision) and different types of stress. For instance have stressful light conditions such as high UV-radiation proved to increase flavonoid and AC production in plants which serve as the plants protection from UV damage. It has also been suggested that nutritional depletion in plants also enhance the transcription of the PAL gene enabling AC synthesis. The other enzymes shown between the steps in the figure (Fig. 8) is cinnamate 4-hydroxylase (C4H), 4-coumarate-CoA ligase (4CL), chalcone synthase (CHS), chalcone isomerase (CHI), flavonone 3-hydroxylase (F3H),

flavonoid 3´-hydroxylase and flavonoid 3´,5´-hydroxylase (F3´H, F3´5´H), flavonol synthase (FLS), leucoanthocyanidin reductase (LAR), dihydroflavonol 4-reductase (DFR),

anthocyanidin synthase (ANS), glycosyltransferase (GT), acyltransferase and malonyltransferase (ACT,MAT) ^{[3, 6, 31]}.

(17)

The anthocyanidin forms are unstable and by glycosylation they become anthocyanins and are stabilized. There are a great number of sugar conjugations possible but the most common are:

glucose, galactose, rhamnose, arabinose and xylase. Glycosylation of the anthocyanidin is what forms the AC since it is where the sugar group is added. If acylation or malonylation occur, it happens after the glycosylation. Anthocyanins are water soluble pigments caused by the presence of sugar. In case the sugar moiety is hydrolyzed or lost, the solubility will decrease and the aglycone will be destabilized or degraded. Acylation and malonylation increase the water solubility of the molecule ^{[3, 6, 33]}.

Keppler et al. employed a new in vitro model system, to investigate the question of deglycosylation, ring scission, and other bacterial degradation pathways of anthocyanins.

They used a naturally composed gut flora directly isolated from the caecum of freshly slaughtered pigs, excluding any aerobic atmosphere. Their results clearly show that

anthocyanidin glycosides were hydrolyzed extensively by the intestinal micro flora depending on the sugar moiety. After cleavage of the protective 3-glycosidic linkage, the released aglycones are very unstable under physiological conditions in the intestine at neutral pH and degrade spontaneously into phenolic acids and aldehydes. Anthocyanidins were released by hydrolysis of ACs. The decay of the aglycone (anthocyanidin) cyanidin led to protocatechuic acid (PCA), peonidin degraded into vanillic acid (VA), and syringic acid (SA) was detected as the degradation product of malvidin. There was also O-demethylation, with the formation of two other metabolites; SA was O-demethylated into gallic acid (GA), and VA converted to PCA. The cleavage of methyl groups and liberation of free hydroxyl groups modulate the antioxidant properties of these phenolic compounds. It is believed that, because of their higher chemical and microbial stability, phenolic acids (hydroxy-benzoic acids) identified as AC metabolites, might be mainly responsible for the observed antioxidant activities and other physiological effects in vivo and not just the anthocyanins themselves ^[11].

2.1.2 Biological effects

Studies show that anthocyanins have substantial bioactivity, including antioxidant activity, and therefore may have beneficial effects on human health ^{[11, 21]}. Dietary ACs’ presumed ability to influence biological systems could be partly due to their characteristic ability to form complexes with macromolecules, combined with their polyphenolic nature. ACs is showed to have a role as scavenger of harmful free radical which cause oxidative damage to nucleic acids, proteins and lipids, in addition to have potential interaction with biological systems. ACs have been shown to have reactivity towards both reactive oxygen and nitrogen in vitro [7, 14, 25]. Reports of health beneficial effects have been shown in humans and also experimental animals. The effects of ACs are many and diverse, ranging from antioxidant behavior to cancer prevention and inhibition. Vision improvement is one function of ACs where they have a regenerating effect of rhodopsin which is a pigment of the retina for formation of photoreceptor cells. Apoptotic effects (programmed cell death) of

anthocyanidins and ACs in cancer cells have been reported both in vitro and in vivo, in addition to tumor decrease, and inhibition of tumorigenesis in rats. Anticarcinogenic effects altering metabolic activation of carcinogens are also reported [8, 11, 14, 22, 34, 35]. There are also studies suggesting that ACs have protective and preventive effects on cardiovascular diseases and atherosclerosis. Atherosclerosis is a condition causing chronic inflammatory response due to thickening of the arterial wall as the result of for instance cholesterol and other fatty

materials. Other functions of the health beneficial effects of ACs are suggested to be anti-

(18)

inflammatory effects, reduced risk of stroke and diabetes, anti-obesity effects and anti-

mutagenic action against various mutagens. In addition to this they are also believed to protect against DNA damage and prevent low-density lipoprotein oxidation and inhibition of platelet aggregation [10, 35-37].

As ACs play a natural part in the daily diet of humans, these discoveries are of high interest and importance to pursue. Knowledge of ACs for prevention of chronic diseases is of great importance for humans as ACs are easily available as food and beverages. To be able to study these effects further also other aspects of these phenolic compounds are needed to research.

The bioavailability and metabolism of the anthocyanins are necessary to study in order to understand more of the biological effects of ACs due to dietary intake.

2.1.3 Bioavailability

To begin to understand the health benefits caused by anthocyanins and the extent of these, there is a need to understand more concerning the bioavailability of ACs in humans. In order to exert an effect in vivo, a dietary compound has to reach tissues, in the native or

metabolized form, in a dose sufficient to yield biological effects ^[12]. To achieve an effect in a target organ or tissue (except the GI tract), bioactive components must be bioavailable.

Evaluating the health benefits of ACs in humans, the bioavailability, including the absorption, distribution, metabolism and excretion, must be known ^{[21, 36]}. ACs have two absorbance maxima, 279-280 nm and 510-540 nm ^[5], a fact which is utilized when trying to detect these compounds. The most common naturally occurring ACs are the following: 3-O-glucosides or 3, 5-di-O-glucosides of cyanidin, delphinidin, peonidin, petunidin and malvidin. Several studies have shown that ACs are absorbed as glycosides in humans, and in experimental animals such as rats [1, 5, 21, 22, 37, 38]. The intact glycosidic forms have indeed been recovered in plasma and urine after oral administration, with one drawback ^[4]. The bioavailability was thought to be very low and considering the metabolism was not, and is still not, fully

understood, the phenolic acid metabolites were not considered at that point. The recovery, of the formerly only known, metabolites (glucuronidated and methylated compounds) in urine and plasma has been found to be below 1 %, whereas up to 99 % if the ingested anthocyanins is eliminated in the gut ^[12].

The supposed low bioavailability of anthocyanins contradicted the explanation of the antioxidant properties of AC rich food as a major health benefit contributor. The results of low absorption values compared to the ingested dosage in humans and animals led to the indication of an extensive biotransformation of the ACs after oral ingestion and absorption. In a study, protocatechuic acid (PCA) was retrieved as the main metabolite of cyanidin

glycosides (CyG). These results showed consistency with former studies, recovering less than 1 % methylated or glucuronidated AC in blood and urine. PCA, however, was found as the main metabolite in serum and it accounted for 44 % of ingested CyG in the following 6 hour period after ingestion. No PCA was recovered in urine, and the native AC and glucuronidated and methylated compounds seemed to be most common. In other studies, PCA has been found in trace amounts in both plasma and urine, unsuspectingly of the source. One study found PCA in concentrations 8-fold higher than that of the native AC itself. In another study, however, PCA was suggested not to be the metabolite of CyG. Despite these discrepancies, PCA was reported as an AC metabolite, and gave way for the possibility of other phenolic acids as metabolites [11-14, 21].

(19)

2.2 Chromatography

Chromatography is a common name for several separation techniques in chemical analysis.

Even though there are different forms of chromatography, they are all based on the principle that the compounds separated are dispersed between two phases; one of them is mobile while the other is stationary. Depending on the method, the mobile phase can either be a liquid or a gas, inspiring the names liquid- (LC) and gas chromatography (GC).

Chromatography can separate gases and volatile substances by GC, non-volatile chemicals and materials of high molecular weight by LC, and also by the inexpensive planar thin layer chromatography (TLC). GC and LC are both column chromatography methods.

In one single process, chromatography can separate a mixture into its individual components, and with an appropriate detector connected identify and quantify each compound. There are of course some restrictions and a need for sample preparation work before an analysis. A column chromatography system consists of the following main components: Mobile phase, pump, column, detector and data treatment (Fig. 9).

Figure 9. Column chromatography system

The selection of the right combination of several separation factors is crucial to the separation of a mixture of compounds:

- Stationary and mobile phase - Length and diameter of column - Mobile phase velocity

- Sample size

The properties of a compound determine the velocity of which the compound moves through the column. The differences between the diverse compounds in a sample make them migrate at different velocities. It is the compounds equilibrium distribution between the two phases that determine the migration velocity. Without this difference there would be no separation.

A sufficient separation is the goal when using chromatography. Resolution (R) is a quantitative measurement for separation. When R has the value of 1, there is 2 % overlap between two peaks. A higher value than 1 indicates a better separation, and a lower value means poorer separation.

(20)

In order to detect the compounds separated with chromatography, a suitable detector is needed. There are several detector possibilities, depending on chromatographic technique and compounds of interest ^[18].

2.2.1 Gas chromatography

Gas chromatography use gas as its mobile phase which is usually called the carrier gas and is stored in a high pressure cylinder. To be able to analyze compounds with GC, the compounds need to be volatile and stable at the temperatures employed by the system. There are two kinds of gas chromatography:

- Gas-solid chromatography (GSC) - Gas-liquid chromatography (GLC)

GLC is the chromatography type used for most GC analyses. There are several nonvolatile compounds that can be separated with both GC and LC but these need to be derivatized before GC separation. The capillary columns in GC have as a rule far higher separation efficiencies than LC columns. Consequently, GC can handle multi-component mixtures more easily.

The purpose of the carrier gas in the chromatographic system is to transport volatile substances through the column. The most common gases used are helium, nitrogen and hydrogen, which are all inert and will not react with the sample or the stationary phase.

Most samples analyzed with GC are fluids which are injected with a syringe often in volumes of 1 µl or less. There are three types of columns for GC; preparative, analytical and capillary

[18].

2.2.2 Detectors

For GC there are several detectors which can be employed, and these can all be classified within two main types:

1) Concentration sensitive detectors 2) Mass sensitive detectors

More than 15 years ago, some of the most commonly used detectors were the thermal conductivity detector (TCD) and the flame ionization detector (FID), type 1 and 2

respectively. During the last two decades mass spectrometry (MS) has increased its impact, and is today the most commonly used detector in combination with gas chromatography (GCMS) ^[18].

2.2.3 Mass spectrometry

A mass spectrometer (MS) is a mass selective detector and fits in under type 2 of GC

detectors.A mass spectrometer consists of an ion source, a mass analyzer and a detector. One of the most frequently used ion sources are electron ionization (EI). When the separated

(21)

compounds enter the EI ion source, a bombardment of electrons hit the molecules resulting in formation of charged molecules and charged fragments (Fig. 10). The charged molecules and/or fragments enter the mass analyzer by the force of electrically charged lenses. Both negatively and positively charged ions are formed in the ion source. However, it is not common to detect negative ions due to the increased sensitivity for positive ionization mode (EI+). Fragmentation is the result of the excess energy in the molecule [18, 39, 40].

Figure 10. Illustration of EI reaction creating charged molecules and ion fragments Several different types of ionization and various mass analyzers can be chosen. The selection of these components is dependent on the use of the MS. The purpose of MS is to separate the ions based on their mass-to-charge ratios (m/z) and to detect these qualitatively and

quantitatively ^{[18, 40]}.

Apart from EI, there are other common ionization methods in the gas phase, such as ^[40]: - chemical ionization (CI)

- field ionization (FI) - photo ionization (PI)

- multi photon ionization (MPI)

Mass analyzers differ with regard to how m/z values are separated, and can be used alone or in combination (MSMS and MSⁿ). The most commonly used mass analyzers are [18, 39, 40]:

- Magnet instrument

Ions with different m/z values are separated in a magnetic field by varying the field.

- Quadrupole instrument

The analyzer consists of four parallel circular rods. Ions are separated in the quadrupole based on the stability of their trajectories in the oscillating electric fields that are applied to the rods. In GCMS combination, mainly quadrupole instruments are used because their mass area covers the molecular weights for the compounds which can be applied to GC.

- Ion trap instrument

The separation happens by a corresponding principle as in the quadrupole instrument. In contrast to a regular quadrupole instrument, the ion source and the analyzer is combined in one common unit.

- Time-of-flight (TOF) instrument

Often used in combination with HPLC but not for GC. Ions accelerate in pulses before they are led through a drift tube where they receive the same kinetic energy and move through the tube with a velocity equivalent to the square root of their mass.

(22)

- Ion cyclotron resonance

The ions are accelerated in a cyclotron and separation occur based on cyclotron frequency of the ions in a fixed magnetic field.

Using a mass spectrometer as detector enables structural information, and the possibility of identification, of the analytes in addition to quantification. There are two modes one can obtain mass spectra with; scan mode or single ion monitoring (SIM) mode. Scan is usually used for identification and SIM for quantification. Continuously scanning the magnetic field exponentially downward from high to low mass keeping the accelerate voltage constant is called “scan” mode. Scan is typically chosen to obtain a mass spectrum of the whole m/z range, and produces mass spectral peaks of constant width. Peak width is inversely

proportional to the resolving power, which is the ability of the MS to separate ions of adjacent mass number. The mass spectrum is a histogram plotted as relative intensity of ions as a function of m/z values.

When components are known or suspected to be found, the full spectrum is not necessary to detect. The MS instrument can also be programmed to detect only one or a few masses, this is called SIM mode. SIM mode requires that the spectrum of the target compounds are known so that the most characteristic masses can be chosen; selecting the right masses for SIM

improves the sensitivity. When masses are chosen for SIM detection, a qualifier (identifier) and quantifier ion is chosen. The identifier ion should ideally be unique for a compound, and the quantifier ion should be the most prominent peak in the mass spectra with the highest intensity compared to the other compound peaks. These criteria ought to be followed to achieve a good foundation for a SIM method ^{[18, 40]}.

2.2.3.1 Typical fragmentation of benzoic acid trimethylsilyl derivatives by electron ionization The combination of GC and MS (or MSMS) is a powerful analytical technique for

identification and quantification of compounds in complex mixtures. By coupling GC directly to the MS, the carrier gas containing the analytes enters directly into the ion source. Ionization can occur since the analytes are volatile and in gas phase.

Sample preparation prior to GCMS analysis of compounds such as BAs involves derivatization to make the compounds compatible with the instrument. A common

derivatization method is silylation creating trimethylsilyl derivatives (TMS) of the compounds in question. By the use of EI as ionization mode, there are several typical ion fragments found in the mass spectra (Table 4).

Table 4. Common ions/ion fragments found in mass spectra of BA TMS derivatives Ions/ion fragments Explanation

[M]⁺ Molecular ion (positive charge) [M‐15]⁺ Loss of a methyl group via α‐cleavage [M‐30]⁺ Loss of a formaldehyde molecule [M‐59]⁺ Subsequent loss of CO2 after rearrangement [M‐89]⁺ Loss of trimethylsiloxyl (TMSO) [M‐177]⁺ Loss of TMSO‐Si(CH₃)₄

(23)

The molecular ion is usually a predominant peak in all BA mass spectra. For TMS esters generation of the fragments [M-15]⁺ and [M-59]⁺ are common. The fragment [M-30]⁺ is produced by derivatives possessing a methoxy group on the phenyl ring. [M-89]⁺ is an established fragmentation pathway for carboxylic acids, because the acylium cation formed is a stable species. The fragment [M-177]⁺ is most common for derivatives with two or more TMS groups attached in adjacent positions ^{[24, 41]}.

Other than the m/z values for the molecular ions and ion fragments, the fragment m/z 73 is commonly observed as the base peak in mass spectra for TMS derivatives, representing the TMS group. The fragment m/z 147 can also be found in such mass spectra, representing the structure [(CH3)2Si=O-Si(CH3)3]⁺ which means there is two or more TMS groups present in the molecule ^[24].

2.2.4 Internal standard

An internal standard (IS) is a substance added to a sample in a known quantity to make it possible to correct certain factors. The IS is an essential component in developing a robust analytical method, and its role is to function as a correction for:

- loss during sample preparation - differences in injected volume

- differences during a chromatographic analysis - change in response factor

During sample preparation there can be loss of sample volume during several steps, for instance by use of pipettes and evaporation of solvent. The internal standard has to be added to the sample solution before the preparation starts.

More than one internal standard may be added to the sample, in case there are multiple substances with different chemical properties which are supposed to be detected in a sample.

In order to quantify compounds in a sample, a standard calibration curve is needed. A

standard curve is prepared by an analysis of known amounts of the substances and the internal standards. The area ratio of substance/IS or peak height ratio of substance/IS is plotted against the concentration ratio^[18].

Equation for a linear calibration curve is:

x = (y-b) / a (1)

Concentration of target compounds are found by this equation:

Conc. compound = (Conc. IS x Compound area) / IS area (2)

There are several criteria that needs to be filled when selecting an internal standard ^[18]: - it has to be separated from the rest of the substances in a given sample, either by time

in the chromatogram or by mass in the mass spectra

(24)

- the retention time must be similar to the chosen substance that is going to be quantified

- similar chemical properties as the chosen substance, which is especially important throughout the sample preparation step such as extraction and derivatization - it must not be present in the sample

- it has to be stable and not react with anything in the sample (or the stationary/mobile phase)

- it needs to be of high purity

A good choice in internal standard is therefore a substance as similar to the substance you want to quantify as possible. One possibility is to utilize an isotopically labeled version of the target substance. Because of the abundance in hydrogen atoms in organic molecules,

deuterium is generally the preferred isotope to use to make isotopically labeled internal standards. A deuterated IS is nearly identical to the substance itself except for its molecular weight which will increase with +1 for each deuterium attached to the substance. Ideally the deuterated IS has the same recovery, very close retention time and the same ionization response, as the unlabeled substance ^[42].

2.3 Sample preparation

Sample preparation is performed before analysis to make a sample suitable for the selected analytical technique. However, before the sample preparation methods can be done, the sample matrix often needs to be somewhat adjusted. Before extraction, a sample might need to be adjusted to the right pH value for the compounds to be able to move to the right phase.

Often, and especially in biological matrixes, compounds are conjugated with highly polar molecules, such as glucuronic acid or sulfuric acid. When the total amount of a compound is required, deconjugation by hydrolysis might be a necessary pretreatment step [10, 15, 18]. Sample preparation is used to purify and concentrate the compounds of interest prior to analysis. The purification usually includes separation of target compounds from the matrix and transference to a solvent which is compatible with the analytical method.

Sample preparation is usually necessary because of the following reasons ^[18]:

- the sample matrix is incompatible with the separation method, and would hamper good analytical quality

- the compounds of interest cannot be analyzed with a satisfactory result because of interference of other compounds in the sample

- the concentration of the compounds are below the instrumental limits of detection The most frequent used sample preparation methods are two extraction techniques ^[18]:

- liquid-liquid extraction (LLE) - solid phase extraction (SPE)

In some cases a simple extraction method for sample preparation is not sufficient prior to analysis of the compounds. Derivatization is chemical modification of a compound, and is often used to increase volatility prior to GC analysis or improve LC detection limits by addition of an appropriate chromophore. The derivatization process chemically modifies one

(25)

compound to produce a new compound which has the suitable properties for the analysis ^[43]. In high pressure liquid chromatography (HPLC) derivatization is not a necessity but it can be utilized to enhance sensitivity and sometimes also affect the selectivity and solubility. For GC, which is the chromatography method used in this thesis, derivatization is used to make compounds more volatile and less polar.

2.3.1 Liquid-liquid extraction (LLE)

Liquid-liquid extraction is an extraction method based on how compounds distributes in two immiscible liquids. Usually, one phase is an aqueous hydrophilic solution, and the other a hydrophobic organic solvent. The sample is the hydrophilic solution and the solvent added needs to have the right properties so that the compounds of interest will transfer over to this phase. To efficiently extract the desired compounds from one phase to another, a sequence of two or three extractions with the chosen solvent is performed ^{[18, 44]}.

The dispersion of a compound in two immiscible liquids can be well described by the equilibrium constant (K) for the given compound at a fixed temperature, ion strength and pH

[18]:

KLLE = [I]L1 / [I]L2 (3)

where [I]L1 and [I]L2 is the two immiscible liquids. If the characteristics of one solvent change the equilibrium constant K changes likewise. To extract a compound from one phase to another the K value is optimized by the choice of solvent, pH and ion strength.

The density of a solvent should not have a higher value than water for the practical benefit of an extraction. It should also be noted that several extractions with smaller volumes yields higher efficiency than extracting once with an equivalent large volume. To concentrate a sample after extraction, the sample is often evaporated, calling for a volatile solvent to be employed for extraction ^{[18, 44]}.

2.3.2 Solid phase extraction (SPE)

In contrast to LLE, the compound is in the case of solid phase extraction not transferred to another liquid phase but to a solid surface, a sorbent. SPE is used with the same purpose as LLE, to isolate, purify and concentrate compounds in liquids^{[18, 45]}.

Equilibrium is formed when the compound is distributed between the liquid sample and the sorbent, either by adsorption to the surface or penetration of the outer layers of the molecules on that surface. An equilibrium constant can also be described for SPE as for LLE:

KSPE = [I]S1/[I]L1 (4)

However, if the dispersion process occurs in a SPE column packed with a sorbent, which is usually the case, a different approach to the equilibrium constant is needed and a different equation applies:

k = 1/(1-k’) or: V0/VR (5)

(26)

where V0 and VR is the void volume in the column and the retention volume respectively ^[45]. If a compound is to be isolated on the surface of a sorbent, the surface needs to have chemical groups which will have a stronger interaction with the compound than the interaction between the compound and the liquid.

There are different packing materials and extraction principles with SPE, the most common are [18, 45, 46]:

- Reversed phase extraction

Used to extract non-polar or weakly polar compounds. It separates analytes based on their polarity, retaining compounds by hydrophobic interactions. The extraction is based on interaction between the carbon-hydrogen bonds in the compound and the carbon-hydrogen bonds on the sorbent. It involves a polar or moderately polar sample matrix and a nonpolar stationary phase (solid phase).

- Normal phase extraction

Used to extract polar compounds from non-aqueous solutions. It involves a nonpolar liquid phase (sample matrix) and a polar solid phase. The principle is based on polar interactions between the compounds and sorbent such as hydrogen bonds, dipole-dipole interaction and other interactions caused by positive and negative charge differences. Normal phase extraction retains compounds based on hydrophilic interaction.

- Ion exchange

Used to isolate ionized compounds from aqueous solutions. The extraction is based on the ionic interactions between sorbent. Anion exchange is based on positively charged functional groups interacting and retaining negatively charged anions (acids), while cation exchange has negatively charged functional groups interacting and retaining positively charged cations (bases).

Silica based sorbents are perhaps the most commonly used sorbent for reversed phase, normal phase and ion exchange columns. Its characteristics inhibit swelling and shrinking in various solvents resulting in faster equilibration in new solvents. The surface chemistry of silica based packing material is the presence of hydroxide groups (silanols), and it is an inorganic polymer with the general formula (SiO2)x. Pore size is usually around 60 Angstrom (Å), prohibiting large molecules from entering the pores, however the pore size ranges from 50-500 Å^{[18, 45]}. Polymer based sorbents are organic polymer materials opposed to silica sorbents which are inorganic. An organic polymer sorbent eliminates problems of highly active sites found on silica and other oxides. The organic compound is usually a polymerization of styrene or methyl methacrylate. To make the polymer useful for SPE, it has to be crosslinked with another compound; typical crosslink agents are divinylbenzene and ethylene dimethacrylate.

The pH range for polymer based packing materials is 1-14 because of short interaction time between sorbent and solvent^[45].

The solid phase extraction is carried out by 4 steps: conditioning of the column, application of sample, washing of column and finally elution of the target compounds. The condition step is performed to activate the column sorbent and preparing it for the sample. If the sorbent is not conditioned before sample application, the functional groups will be compressed on the

(27)

surface and the active surface area will be very little. This is however more relevant for silica based columns than for the polymer based columns. Thus, the goal of conditioning is to increase the surface area. Thereafter, washing of the column is performed by a solution of similar composition as the sample. Washing of the column is typically done to get rid of the conditioning solvents. When the sample is applied to the column, the target compounds will interact with the packing material and become retained while other molecules and impurities pass through. By adding an additional wash step, components with weak interaction to the sorbent gets washed out. Target compounds can be eluted from the column by applying an elution solvent which should not be strong enough to elute pollutants. It is also of great importance when using silica based columns to not let the column dry between the

conditioning, washing and sample application steps, to prevent packing material to collapse thus spoiling the extraction process ^{[18, 45]}.

2.3.3 Derivatization

To make compounds compatible with chromatographic analysis, derivatization is often needed. Derivatization is especially performed prior to GC, to convert water soluble and temperature labile compounds suited for GC analysis. By derivatizing compounds certain properties will increase:

- volatility - detectability - stability

By optimizing the derivatization method to increase these properties, the peak shape, intensity and resolution will be better.

There are common types of derivatization prior to GC-analysis:

- silylation - alkylation - acylation

Silylation is the most commonly used derivatization method in GC, allowing a wide variety of compounds to be silylated and a large number of reagents are available. The derivatization method used in this thesis is silylation. The silyl derivatives are more volatile and more thermally stable than the original compounds. Trimethylsilyl (TMS) groups will replace active hydrogens on the compounds treated with the derivatization reagent. Trace alcohol will also react with the derivatization reagent leading to insufficient amounts of derivatization reagent to fully derivatize the target compounds. This fact makes it crucial for the end result to dry the sample and solvent adequately before derivatization. The reagents are sensitive to moisture and because of its derivatization mechanism only aprotic organic solvents can be used. The ranking of functional groups after silyl accepting properties are: alcohols > phenols

> carboxylic acids > amines > amides [18, 43, 47].

There are several silylating reagents, including N-trimethylsilylimidazole (TMSI/TMSIM), N,O-bis(trimethylsilyl) trifluoroacetamide (BSTFA), N,O-bis(trimethylsilyl) acetamide (BSA), trimethylchlorosilane (TMCS) and hexamethyldisilazane (HMDS). The reagents are ranked in that order respectively with regard to their silyl donor properties ^{[18, 43]}. BSTFA and

(28)

BSTFA+TMCS are the preferred reagents for trimethylsilylation of alcohols, alkaloids, amines, biogenic amines, carboxylic acids, phenols, and steroids. The use of BSTFA and BSTFA+TMCS as the most common derivatization reagents of phenolic acids prior to GCMS analysis seem to correspond to literature [19, 24, 25, 41, 48, 49]. TMCS is typically mixed with other silylating reagents to increase their reactivity and work as a catalyst. TMCS is rarely used alone in analytical applications ^{[50, 51]}.

The mechanism of BSTFA (and TMCS) as silylating reagent, where an active hydrogen is replaced by an alkylsilyl group - TMS, is shown in Fig. 11 ^{[47, 51]}.

Figure 11. Silylation mechanism of reagents BSTFA and TMCS

Certain chemicals, other than TMCS, can be used to enhance the derivatization result when added to the derivatization reaction. Examples are chemicals such as the solvents pyridine and acetonitrile which both have positive effects as catalysts. A catalyst is used to increase the reactivity of the reagent ^[24]. Pyridine is also used to catch HCl formed from the derivatization processes where chloride is part of the reagent. The free nitrogen on the pyridine molecule reacts with the chloride ion.

2.3.4 Solid phase analytical derivatization (SPAD)

SPAD is an example of the combination of two sample preparation methods in one. The target compounds are adsorbed to the packing material, followed by addition of derivatization reagent, and finally elution of the derivatized compound. By merging SPE and derivatization, time will be saved as both steps are time-consuming.

(29)

3. Materials and methods

3.1 Chemicals

Acetic acid, acetone, acetonitrile (AcN), ethyl acetate (EtAc), hydrochloric acid (HCl) fuming 37 %, methanol (MeOH), sodium sulphate (anhydrous) and tert-butyl methyl ether (MTBE) were purchased from MERCK (Darmstadt, Germany). β-glucuronidase with 5 % sulfatase activity (Type HP-2 from Helix pomatia), pyridine (anhydrous) and sodium acetate (anhydrous) were bought from Sigma-Aldrich (Steinheim, Germany). The derivatization reagents N,O-bis(trimethylsilyl) trifluoroacetamide (BSTFA) and BSTFA +

trimethylchlorosilane (TMCS) 99:1 (Sylon BFT), were obtained from Supelco (Sigma- Aldrich). All chemicals are of HPLC purity or better.

3.1.1 Solutions

Abbreviations and purities for benzoic acids (BAs) are spelt out in Table 5. Stock solutions were prepared for all six benzoic acids in addition to HMBA which was used as an internal standard for the GC system. Stock solutions of single compounds GA, PCA, VA, DHMBA, SA, HBA and HMBA were made by dissolving BAs in 1 mL 100 % MeOH with

concentrations ranging from 5.79 to 13.33 mg/mL. Concentrations were calculated by weighing both BA and solvent on an analytical balance (accuracy of ± 0.01 mg), as well as considering BA purity. Deuterated internal standards of GA and HBA were prepared in the same way, with concentrations 8.86 and 11.45 mg/mL respectively. All user and standard solutions were prepared from these stock solutions. Accurate concentrations for all stock solutions are found in Appendix 1.

Table 5. Purity and distributor of the benzoic acids and the internal standards

Chemical name Abbreviation Common name Purity (%) Distributor

3,4,5‐trihydroxybenzoic acid GA Gallic acid 100 Sigma

3,4‐dihydroxybenzoic acid PCA Protocatechuic acid 100 Fluka 4‐hydroxy‐3‐methoxybenzoic acid VA Vanillic acid 99.40 Fluka

3,4‐dihydroxy‐5‐methoxybenzoic acid DHMBA ‐ 99 Fluorochem

4‐hydroxy‐3,5‐methoxybenzoic acid SA Syringic acid 99.25 Fluka

4‐hydroxybenzoic acid HBA ‐ 99 Aldrich

4‐(hydroxymethyl)benzoic acid HMBA ‐ 100 Aldrich

3,4,5‐trihydroxybenzoic‐2,6‐d2 acid d‐GA Deuterated gallic acid 99.30 CDN Isotopes 4‐hydroxybenzoic‐2,3,5,6‐d4 acid d‐HBA ‐ 98.80 CDN Isotopes

PBS buffer was prepared (0.1 M, pH 7.2), using BupH^TM Phosphate Buffered Saline Pack from Pierce, containing 0.1 M sodium phosphate and 0.15 M sodium chloride.

The enzyme solution was prepared by mixing 100 µL β-glucuronidase with 5 % sulphatase activity (H-2, Helix pomatia, 131 700 units/mL) with 2900 µL sodium acetate buffer (0.4 M,

(30)

pH 4.95).

3.2 Standard calibration curve

Standard solutions, diluted with MTBE, were prepared from the stock solutions made of all BA and IS, to make a standard calibration curve (Appendix 2). Ten calibration standards, containing a mix of six BAs, two deuterated quantitation internal standards and one GCMS internal standard were prepared in nine concentrations ranging from 5.29 to 1654.52 µg/L, in addition to a calibration blank (Table of accurate concentrations found in Appendix 3). The internal standard were added to all standards with constant concentration; 197.06, 183.53 and 172.32 µg/L for HMBA, d-HBA and d-GA in that order.

The standard calibration curves used throughout the analyses made in this thesis were linear (eq. 1 in 2.3.5). BA concentrations were found by the principle of eq. 2 in section 2.3.5:

Conc. BA = (Conc. BA IS x BA area) / BA IS area (6)

Several calibration curves were prepared throughout the analysis work because fresh standard solutions were made every second month to obtain reliable and reproducible results. Three internal standards were used in the analyses, two were the deuterated substances

corresponding to the unlabeled substances GA and HBA; d-GA and d-HBA respectively, and the third was HMBA. The two isotopically labeled internal standards were used as BA IS and labeled IS for the standard calibration curve which the BAs are quantified by. The third IS, HMBA, called the GCMS IS was used to keep track of the instrument stability. If the GCMS IS concentration changes noticeably between analyses, the instrument sensibility is changing and gives an indication that a new calibration curve should be made.

3.3 Sampling

All samples analyzed for the utilization part of this thesis has been stored at - 20ºC or - 72ºC at different periods of time before analysis. Samples used for method development has either been stored at -20ºC or collected shortly before sample preparation.

3.3.1 Urine samples

Urine samples were given by volunteers (n=13) at different periods of time. An important aspect of analyzing urine samples is that they need to be comparable. To be able to compare BA concentrations in different urine samples, either the total urinary volume within each sampling period or the creatinine level was measured.

The urine sampling can be divided into three categories:

- Method development

Spot urine from healthy volunteers (n=5) was used as matrix for development and evaluation of the best sample preparation method. Urine used from method development was mainly collected shortly before preparation. For method

(31)

development considering deconjugation samples stored at – 20 ºC was utilized because of known levels of creatinine.

- Oral administration of MEDOX®

For samples that were to be analyzed after oral administration of Medox® capsules, two kinds of sampling were employed:

1) Spot urine sampling

Ten healthy volunteers, ranging from 24 to 60 years of age, were orally administered 6 capsules of Medox®. Before and two hours after ingestion of the capsules, spot urine samples were collected. The volunteers did not follow any diet restrictions. The urine samples from this experiment are from 2007 and have been stored at -20 ºC.

2) 6 hour sampling

Two healthy female volunteers were orally administered 8 Medox® capsules. Urine were sampled before and 1, 2, 3 and 6 hours after administration. Total urinary volume was measured for each time interval, as well as creatinine levels. No diet restrictions were followed by the volunteers. All collected urine samples were stored at -20 ºC.

- Oral administration of delphinidin-3-O-β-glucopyranoside

One healthy male volunteer was orally administered 500 mg delphinidin-3-O-β- glucopyranoisde powder, swallowed with water. One urine sample was collected before ingestion of the delphinidin powder. Urine samples were also collected at each of the following 8 hours after administration. The volunteer followed a low

anthocyanin diet for 48 hours prior to the administration and during the sampling period. All samples were acidified (1 mL 6 M HCl to 25 mL urine), aliquoted and stored at -72 ºC. The samples from this delphinidin experiment have been stored since 2007.

3.3.2 Plasma samples

Plasma (EDTA) samples were collected from healthy volunteers (n=4) in a similar manner as the urine samples. All blood samples were collected in two 10 ml EDTA vials, and 15 minutes after collection they were centrifuged for 10 minutes at 2000 g and 4 ºC to separate the plasma from the remaining blood cells. Plasma from both vials was mixed to a

homogenous sample.

The sampling can be divided into the same three categories as the urine samples were:

- Method development

Plasma samples used for method development had been stored at -20 ºC.

- Oral administration of MEDOX®

For plasma samples, one kind of sampling considering oral administration of Medox®

was applied. Blood was collected identically to the 6 hour sampling type for urine, with the same two female volunteers administered 8 Medox® capsules. Samples were drawn before ingestion and then 1, 2, 3 and 6 hours following administration. All plasma samples were stored at - 20ºC.

(32)

- Oral administration of delphinidin-3-O-β-glucopyranoside

The same healthy male volunteer as for urine sampling was orally administered 500 mg delphinidin-3-O-β-glucopyranoside powder swallowed with water. A blood

sample was drawn before the ingestion of delphinidin. Subsequent blood samples were collected after time intervals: 15, 30, 45, 60 and 90 minutes, and 2, 3, 4, 5, 6, 7 and 8 hours. Plasma was acidified (40 µL 6 M HCl to 1000 µL plasma) and stored as 1 mL aliquots at – 72 ºC. These samples have been stored since 2007.

3.4 Sample pretreatment

Pretreatment of samples before preparation step included pH adjustments and enzymatic reactions. Both urine and plasma samples were adjusted to pH 2 by addition of 1 M HCl before extraction by LLE, SPE and SPAD.

3.4.1 Deconjugation

Some samples were deconjugated before sample preparation; this step was performed before acidification of samples because of optimal enzyme reaction at pH levels 4.5 – 5.

Deconjugation of a sample involved adding 300 µL enzyme solution to 500 µL urine or plasma. The composition of the enzymatic mixture was 10 µL enzyme + 290 µL sodium acetate buffer (see 3.1.1). The mix was incubated for 2 hours at 37 ºC.

3.5 Sample preparation

All frozen samples were thawed in the refrigerator and subsequently went through a

pretreatment step before sample preparation. The sample preparation consists of an extraction method and derivatization. Between the extraction and derivatization step, the sample was concentrated by evaporation either by nitrogen flow or vacuum centrifugation (except for in SPAD where the two preparation steps are combined).

All three extraction methods (LLE, SPE and SPAD) were compared with the consideration of recovery values, peak resolution and execution.

3.5.1 Derivatization

BA sample extracts and standards were derivatized by the final optimized method as follows:

Sample extracts or standards were derivatized in a mix of 25 % AcN, 10 % BSTFA and 65 % MTBE, by heating at 60 ºC for one hour.

The optimized method described above was achieved by variation of the following parameters: